Article Text


Local anaesthesia for 1221 vitreoretinal procedures
  1. R S B Newsom,
  2. A C Wainwright,
  3. C R Canning
  1. Southampton Eye Unit, Southampton General Hospital and Shackleton Department of Anaesthesia, Southampton General Hospital, Southampton, UK
  1. Mr R S B Newsom, Vitreoretinal Unit, Moorfields Eye Hospital, City Road, London EC1V 2PD, UKrbnewsom{at}


BACKGROUND Vitreoretinal (VR) surgery has been increasingly performed under local anaesthesia (LA) in this unit. The results of an audit monitoring this change are presented.

METHODS Data were collected on 1497 patients including type, volume, and position of the anaesthetic block, type of surgery, complications, and patient reactions.

RESULTS 1221/1479 (82%) patients had LA. They were older than those having general anaesthesia (GA) 63.5 years v 45.9 years; 146 (10.6%) blocks were intracone, 130 (10.6%) peribulbar, and 927 (75.9%) combined. Operations included 436 vitrectomies, 545 retinopexy with or without vitrectomy, and 238 buckling procedures. Some pain was felt by 9.4%, 8.8%, and 19.7% of patients during vitrectomy, retinopexy with or without vitrectomy, and buckling surgery respectively.

CONCLUSION Local anaesthesia for VR surgery is well tolerated by patients, being effective throughout longer and more stimulating ocular surgery.

Statistics from

For many ophthalmic surgeons local anaesthesia (LA) has become the preferred option over general anaesthesia (GA) because of quicker patient rehabilitation and the avoidance of possible complications from general anaesthesia.1-3

Several methods of LA for vitreoretinal (VR) cases have been described, including retrobulbar,4-9 peribulbar,8-11sub-Tenon's,1213 and even topical anaesthesia in some cases.1415 In several countries 85–90% of VR surgery is carried out under LA. However, this is not the case for VR surgery in the UK.4 Reluctance to use LA for VR surgery may stem from the longer, more unpredictable and uncomfortable nature of VR surgery.

In our unit LA has been introduced over the past 6 years and has increased in use from 12% to 96% in January to 1 February 2000. An audit of LA for VR surgery was performed over this transition period and the data presented in this paper are a summary of this experience.


Data were collected prospectively on 1497 consecutive patients undergoing VR surgery, over a 6 year period. All patients were anaesthetised by a single anaesthetist (ACW). The patient's age, previous surgery, type of anaesthetic, anaesthetic agents, and volumes were noted.

Akinesia was assessed on a range of 0–4 (0 no movement, 1 flicker, 2 some movement, 3 moderate movement, 4 full movement). Reaction to the LA block was graded from 1–4 (1 being no pain felt, 2 no comment, 3 moderate discomfort, and 4 severe pain).

Surgery was divided into vitrectomy, retinopexy with or without vitrectomy, and scleral buckling surgery. Patients' reaction to surgery was graded 1–4 (1 minor discomfort—for example, to subconjunctival injection, 2 mild, 3 moderate, 4 severe—would not like LA again). Anaesthetic complications affecting surgery were noted and use of peroperative sedation.

Data were analysed using statview software. Non-parametric data are presented as proportions with percentage and 95% confidence intervals (95% CI). These were analysed using χ2 test. Parametric data were analysed using the unpaired two tailt test. Confidence intervals omitted from the text are shown in Table 1.

Table 1

Demographic data


In all, 1221/1479 (82%) patients had local anaesthesia for VR surgery (Table 1). Operations under LA included 436/1221 (35.7%) vitrectomies, 545/1221 (44.7%) retinopexies with or without vitrectomy, and 240/1221 (19.5%) buckling procedures. LA was used in 203/258 redo procedures (79%; 95% CI 76.5–81.3%).

Patients under 35 years (younger) were more likely to have a general anaesthetic, 87/147 (59.1%: 95% CI 54.0–64.5%), those over 35 years (older) were more likely to have LA, 1161/1332 (87.1: 95% CI 86.1–88.1; p<0.01) (Fig 1). There was no difference in peroperative pain (10.6% v 11.0%) or need for sedation (10.4% v 7.6%) in younger or older patients, respectively. However, younger patients were more likely to feel discomfort during the local block (31.9%v 14.9%; p<0.01) and have more kinesia (17.0% v 6.3%; p<0.01).

Figure 1

Histogram of patients' age versus type of anaesthesia used for vitreoretinal surgery. 59.1% of patients aged <35 years had general anaesthesia (GA); 87% of patients aged >35 years had local anaesthesia (LA).

Of the 1221 LA blocks 164 (13.4%) were intraconal only, 130 (10.6%) were peribulbar only, and 927 (75.9%) were a combination of both. The volumes of LA used for inferotemporal, medial, and top up injections were: 5.3 ml (SD 0.53), 3.14 ml (1.31), 0.7 ml (1.78), respectively. The total volumes given for vitrectomy, retinopexy with or without vitrectomy, and buckling surgery were 7.94 ml (1.6), 8.52 ml (1.8), and 12.58 ml (2.98), respectively (p<0.001). The proportion of patients feeling pain (score 3–4) from the anaesthetic injection was significantly lower when bicarbonate was added, falling from 94/408 (23.0%; 95% CI 21.0–25.1%) to 178/1813 (11.6%; 95% CI 10.33–12.9, p<0.01). More patients having redo surgery following buckling felt pain during the anaesthetic (41/203; 20.14%; 95% CI 16.1–23.1% v 153/1081; 15.03%; 95% CI 13.9–16.1, p<0.05). Peroperative top up anaesthesia was given in 193/1221 (15.8% 95% CI 14.8–16.8%); this rose to 159 /240 (66.2% 95% CI 63.2–69.2%) patients having buckling surgery.

Compete akinesia was noted in 1134 patients (92.8% 95% CI 92.1–93.6), slight twitch in 79 (6.74%: 95% CI 5.8–7.2), moderate kinesia in seven (0.6%: 95% CI 0.4–0.8%), and full movement in one (0.08%: 95% CI 0–0.2%).


Pain scores of 0, 1, 2, 3, 4 were reported by 1085/1221 (88.9%; 95% CI 88.0–89.8%) 69/1221 (5.7%; 95% CI 5.0–6.3%), 57/1221 (4.79%; 95% CI 4.1–5.3%), 7/1221 (0.57% 95% CI 0.36–0.79), and 2/1221 (0.2%; 95% CI 0–0.3%) patients, respectively.

The percentage of patients with any pain during surgery varied with operation and were 17/435 (3.9%; 95% CI 2.9–4.8), 25/543 (4.6%; 95% CI 3.7–5.5), and 24/240 (10.0% 95% CI 8.1–11.9) for vitrectomy, retinopexy with or without vitrectomy, and buckling surgery respectively. Buckling surgery was more painful (p=0.002) and more sedation was required in 28/240 (11.67%, 95% CI 9.6–13.7), compared with vitrectomy 26/436 (5.9%, 95% CI 4.8–7.1), (p=0.016) (Table 1).

There was also no difference in pain between primary (51/1081; 5.0% 95% CI 4.3–5.69) and redo surgery (16/203; 7.9% 95% CI 5.9–9.7%).


Sedation was given to 95/1121 (7.8%, 95% CI 5.0–10.5%) of patients. Sedation was used in 41/545 (5.96%; 95% CI 4.8–7.1%) of vitrectomies, in 26/436 (7.52%; 95% CI 6.3–8.7%) patients with retinopexy with or without vitrectomy and in 28/240 (11.67%; 95% CI 9.6–13.7%) patients needing buckling surgery (p<0.01). Sedation was required in 16/203 (7.9%; 95% CI 6.0–9.8%) patients under going redo surgery, compared with 79/1018 (7.8%; 95% CI 6.9–8.6%) of those undergoing primary surgery (p>0.1). Patients needing sedation were given more LA 9.8 ml (SD 3.03) compared with those not needing sedation 9.05 ml (2.6, p<0.005) (Table 1).


No patient suffered a retrobulbar haemorrhage or ocular perforation in this series: 64/1221 (5.24%; 95% CI 4.6–5.8%) patients had moderate to severe chemosis, none of which prevented surgery taking place, 4/1221 (0.33%; 95% CI 0.2–0.5%) patients had significant bradycardia, 9/1221 (0.74%; 95% CI 0.4–0.9%) had moderate claustrophobia, 1/1221 (0.08%; 95% CI 0–0.016%) severe claustrophobia and was given propofol, and one operation was cancelled (0.08%; 95% CI 0–0.016%) (Table 1).


In this series we have been able to assess the results of LA for VR surgery in 1221 cases given by a single anaesthetist (ACW) and have found it to be a well accepted and useful technique. The high levels of patient satisfaction in this study are similar to the study of Raoet al4 with 93.1% of patients happy to have LA again.

Using a combined peribulbar and intraconal technique, good analgesia and good akinesia were obtained in 87.9% and 98.0% respectively. When bicarbonate was added to the LA solution the proportion of patients with discomfort during anaesthesia fell to 11.6%. These high levels of akinesia and anaesthesia are comparable with other methods of anaesthesia.1216 A top up rate of 32% has been reported with peribulbar and retrobulbar techniques,8 in our series it was 16% rising to 66% in buckling surgery.

Redo operations may be a contraindication to LA. However, in this series we performed 79% of second procedures under LA, a higher proportion of those with previous buckling surgery felt some pain during the LA (11.6% v 10.5%); however, peroperative pain was similar in both groups (7.8%v 5.1%). Midazolam was used in 95 patients; in our experience sedation was useful in controlling claustrophobia (10/1221).

Kirkby et al5commented that a oculocardiac response and retrobulbar haemorrhage may prevent surgery on urgent cases. In this series, no cases of retrobulbar haemorrhage or globe perforation occurred, significant bradycardia was noted in four patients, chemosis occurred in 61 of our patients but this was never sufficient to prevent surgery.

In this department LA was originally used for postoperative analgesia; this led to its use for high risk patients and shorter procedures. Now careful patient selection and counselling have enabled the use of LA for 96% of VR procedures with high rates of surgeon and patient satisfaction.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.